Greg Detre
Wednesday, March 19, 2003
Steve
Larson: maths as existing outside a consistent formal system?
pain as
being an obstacle to other goals
drugs to
keep awake
ideonomy � Dexadrone � Patrick Gunkel
new drug that keeps pilots awake for 36 hours
students taking crushed Ritalin
Jonathan Grover: mefanol??? mefadanol???
sees
Chomsky as a neo-behaviourist in terms of not producing ideas that are powerful
enough to explain linguitstic phenomena
pain
asymbolia � lesion cingulate gyrus
discussing
non-expressed genes:
difficult for evolution to weed them out
backups, residues from other animals
apparently the frugal fish doesn�t have introns
neither do bacteria, apparently
how did the foogoo(sp???) (zebra fish???) get
rid of its non-expressed (introns) genes??? how big is its genome???
he doesn�t
consider there to be a neuroscientific theory of LTM
he doesn�t
think you can get very far bottom-up without some good architectural ideas
about what to look for at the middle-level�
he asks: what�s
a low-resolution representation of 1000 statements in English?
apparently
adults can learn a new language faster than children � if you go on an
intensive 3-month course � learning 50 words a day
the child has to learn the concepts at the same time, whereas adults
already have them
there are certain things (like accents) that adults can�t learn
the adult is spending all day every day learning the language, whereas the child is
presumably only hearing people every so often
harder to do maths in a foreign language (even if you learned it in your
teens) � hmmmm
the numerals don�t change � perhaps it�s
harder to relearn the associations
then what about people who learn a language
with different numeral shapes???
refce: stanislas dehaene, �the number sense � how the mind creates
mathematics� � apparently, argues innateness of the number line (for
example) as a cognitive structure � that sounds very plausible
Minsky to
his daughter � �if you do that, I won�t give you your allowance for 6 weeks� �
�you�ll forget�
when people
are hypnotised and someone�s telling you something, you can edit their memories
� now, you can use hypnosis for clues, but not for evidence
story about a guy who hypnotises and asks his student to regress to 6
years old, handwriting changes etc. � then he tells the student to �pregress�
10 years into the future, and the guy describes what he�s doing, and the case
he�s had that day etc.
flashbulb
memory?
study asking people where they were during 9/11, but they were
completely inaccurate
the problem with asking people about JFK�s death is that they�ve
rehearsed their answer so many times
re Ron
Hubbard and childhood abuse � they can almost never find the room that�s been
described as where it happens
very hard to find reliable memories laid down before 5 or so (infantile
amnesia)
most
reasoning by analogy, rather than formal deduction
after, you
construct a logical story
but he
thinks that we are capable of doing logical reasoning
common
sense knowledge has too many exceptions for deductive to be useful
deductive
logic isn�t something that we�re good at
presumably we have a module that gets better at clean abstract deductive
reasoning over a lifetime
I think he does think that we�re ultimately capable of deductive
reasoning though
children
don�t have a (uniform) sense of time
Piaget
apparently didn�t write 2nd drafts, and is usually better in translation J
refce: read Winston�s thesis
interesting
idea: use legal disclaimers as a source of common sense knowledge J
intelligence
is when you see someone do something you want to be able to do
cs as all
the things that don�t need to be said
the thing
about IQ is that it�s relative to your age � explanation of the Flynn effect �
could it be that if everyone�s maturing slower, their IQs will appear to go
up??? does that work arithmetically???
how do they even compare IQs today with last century
Bettie
Edwards � you can�t draw something you recognise � but you can if you just draw
the spaces in between!
Galway �
Inner tennis � don�t keep your eye on the ball, because by that time it�s too
late
juggling �
start with one ball, look at the top of the parabola, rather than keeping your
eye on the ball � then you learn unconsciously to figure out where it�s going
to land
untrained
reflexes c. 0.2s, fastest athletes, 0.15s � false start in the 100m if you go
before 0.15s
try to hit
an imaginary ball 3 inches beyond the white (that way the extensors haven�t
started to slow your arm down when you reach your actual white-ball
destination)
pianist
playing skips (large arm movements) � each hand covers an octave, with one
octave either side, covering 6 in total, and the remaining 2 which don�t ever
get used
he thinks
there should be a good with 10 great tips for each field � hmmm
they might be good for boostrapping
but most valuable things are hard-earned � Piagetian??? Calvinist J
I�m skeptical about our ability to write down the things we learn
Ed Emberly � drawing tips, �Make a world�
he likes
the idea of teaching vertical skills (that build on top of prior learning) in
schools (where you can generate more and more complexity out of your building
blocks)
last week,
you said that the reason you weren�t troubled by Godel�s theory is that you
consider the brain to be a formal system, but an inconsistent formal system
I don�t understand his answer
are we
really �rational� animals??? what would that mean (if we�re inconsistent)???
doesn�t
that undermine all thinking/philosophising/discussion, since we have no appeal
to consistency as a rational authority???
rationality
as more than just a Difference Engine???
you talk a
lot about intelligence, but that seems like a much more creative, speculative,
forward-looking process than rationality, which is maybe more of a checking,
backwards-looking process�???
why do we make mistakes???
why would we have evolved to be completely rational � difficult/veiled truth is not adaptive???
why don�t
you discuss analogy???
do you think it�s involved in learning???
he does discuss it � where???
why are some concepts more difficult???
perhaps because it�s difficult to find analogies for them???
do you
reckon that we can understand ourselves???
how can our brains represent themselves???
perhaps by abstracting + devolving most of the difficult work to the
simulations???
doesn�t it kind of rest on the naturalist premise that the world is
describable and discoverable by science�???
wrt Godel,
could it be that we have islands of consistency???
what
happens if we put a hard disk in our heads � would it be conscious???
presumably he�d say that if it affects our agents it would be � hang on
though, he couldn�t say that, because there are lots of causal events further
upstream than my senses that ultimately affect my agents that I�m not conscious
of
common sense vs intelligence???